Prediction of the Behavior of a Power System Using Root Cause Failure Analysis

Abstract

The thermo-economic performance of a gas turbine is simulated using a fish bone technique to characterize the major equipment failure causes. Moreover a fault tree analysis and a Pareto technique are implemented to identify the related failure modes, and the percentage and frequency of failures, respectively. A pump 101 and drier 301 belonging to the Tabriz Petrochemical Company are considered for such analysis, which is complemented with a regression method to determine a behavioral model of this equipment over a twenty-year period. Research findings indicate that 81% of major failure factors in production equipment are related to the executive procedures (24%), human error (22%), poor quality of materials and parts (20%), and lack of personnel training (15%)